amor{thetime-of-°ightneutronre°ectometer atsinq/psi€¦ ·...

PRAMANA c© Indian Academy of Sciences Vol. 63, No. 1— journal of July 2004

physics pp. 57–63

AMOR – the time-of-flight neutron reflectometerat SINQ/PSI

MUKUL GUPTA1, T GUTBERLET1, J STAHN1, P KELLER1

and D CLEMENS1,2

1Laboratory for Neutron Scattering, ETHZ & PSI, Paul Scherrer Institute, Villigen,CH-5232, Switzerland2Hahn-Meitner-Institut Berlin, Glienicker Str. 100, D-14109 Berlin, GermanyE-mail: [email protected]; [email protected]

Abstract. The apparatus for multioptional reflectometry (AMOR) at SINQ/PSI is aversatile reflectometer operational in the time-of-flight (TOF) mode (in a wavelengthrange of 0.15 nm < λ < 1.3 nm) as well as in the monochromatic (θ–2θ) mode with bothpolarized and unpolarized neutrons. AMOR is designed to perform reflectometry measure-ments in horizontal sample-plane geometry which allows studying both solid–liquid andliquid–liquid interfaces. A pulsed cold neutron beam from the end position of the neutronguide is produced by a dual-chopper system (side-by-side) having two windows at 180◦ androtatable with a maximum frequency of 200 Hz. In the TOF mode, the chopper frequency,width of the gating window and the chopper–detector distance can be selected indepen-dently providing a wide range of q-resolution (∆ q/q = 1–10%). Remanent FeCoV/Ti :Nsupermirrors are used as polarizer/analyzer with a polarization efficiency of ∼97%. For themonochromatic wavelength mode, a Ni/Ti multilayer is used as a monochromator, giving∼50% reflectivity at a wavelength of 0.47 nm. In the present work, a detailed descriptionof the instrument and setting-up of the polarization option is described. Results fromsome of the recent studies with polarized neutrons and measurements on liquid surfacesare presented.

Keywords. Apparatus for multioptional reflectometry; neutron reflectivity; neutron re-flectometer; time-of-flight; polarized neutrons; liquid interfaces.

PACS Nos 61.12.Ha; 75.25.+z; 83.85.Hf

1. Introduction

In recent years X-ray and neutron reflectometry have been established as non-destructive microscopic probes for the investigation of stratified structures andhidden interfaces. Due to the unique interaction of neutrons with matter, specu-lar neutron reflection probes the atomic and magnetic scattering length densitiesperpendicular to the sample surface. It can be very sensitive to electrochemicalprocesses at interfaces as well as to surface and interface magnetism. Neutron scat-tering length density contrast among the isotopes of an element provides an oppor-tunity for measuring self-interdiffusion in a multilayer stack. Moreover, polymer

57

Mukul Gupta et al

interfaces and biological model membrane systems can be studied using contrastvariation. Growth, wetting, adsorption and adhesion, (self)-interdiffusion, surfacemagnetism, magnetic excitations, thin film superconductivity, dynamics at inter-faces are all research fields applicable to neutron reflectometry.An extremely flexible state of the art TOF reflectometer, AMOR has become

now available at the Swiss neutron source SINQ at PSI [1–3]. AMOR is adaptableto the experimental demands of surface and interface studies in various fields ofresearch. Full polarization analysed neutron reflectivity measurements, measure-ments of liquid samples using a fully automated Langumir trough and solid–liquidsample cell are recent developments at AMOR.

2. Basic instrument concept

The principal set-up of AMOR allows measurements with polarized or unpolarizedneutrons in white beam TOF mode (0.15 nm < λ < 1.3 nm) and optionally inunpolarized monochromatic θ–2θ mode. The latter mode is implemented with athin film Ni/Ti multilayer monochromator consisting of 1500 layers with a bilayerperiod of 5.2 nm.The plane of the sample is kept horizontal in order to allow measurements at

open liquid surfaces, too. The inclination angle and thereby the accessible q-rangeis adjusted by tilting a deflection mirror and/or the sample. A flexible softwarecontrol of the θ–2θ movement around the axes that are not mechanically coupled hasbeen implemented. The standard mode of the instrument is TOF, which has beenopen for user operation since Oct. 2002. An area detector or two single detectortubes can be operated alternatively with new fast detector read-out electronics.Most optical components are riding on an 8 m optical bench so that the chopper–detector distance can be varied in order to give an optimum sample illuminationand resolution (figure 1). In the following sections each of these parameters aredescribed.

3. Chopper system

A cold neutron beam from the end position of the curved neutron guide (m = 2,where m corresponds to the critical edge of Ni) 1RNR17 with a flux of 1.38 ×108 n cm−2 s−1 mA−1 (averaged over the spectrum) hits a double chopper systemdefining the origin of the neutron burst of the instrument. The neutron guide hasa cross-section of 5 × 5 cm2. The spectral distribution obtained from the liquid D2

moderator at SINQ offers a wide usable wavelength band which peaks at λmax = 0.4nm. The two chopper discs have two slits at 180◦. They are phase coupled with amaximum speed of 6000 rpm giving burst rates of 200 Hz. The distance betweenthe chopper and the detector varies between 3.5 and 10 m. The chopper frequency,width of the gating window and the chopper–detector distance can be selectedindependently. Hence the resolution can be optimally tuned to the experimentalneeds in a range of ∆q/q = 1–10%.

58 Pramana – J. Phys., Vol. 63, No. 1, July 2004

AMOR – the time-of-flight neutron reflectometer at SINQ/PSI

Figure 1. Reflectometer AMOR at SINQ/PSI and principal layout.

4. Optics

A frame overlap mirror consisting of a Si based Ni/Ti supermirror (m = 2) posi-tioned behind the chopper system eliminates undesired neutrons with wavelengthslarger than 1.3 nm. Following this position is a station that can be equipped byeither a deflecting mirror of a highly oriented pyrrolytic graphite (HOPG) doublemonochromator to give a monochromatic beam of λ = 0.4735 nm or by a rema-nent polarizing FeCoV/Ti :N supermirror (m = 2.6) that can operate for a broadwavelength band of polarized neutrons. The performance of the polarizing mirroris shown in figure 2.The sample stage to align the sample horizontally (vertical scattering geome-

try) is equipped with two rotation angles and a translation in z. Samples of sizeup to 150× 500 mm2 can be measured at ambient atmosphere when placed on astandard sample mount. Behind the sample stage, a reflecting mirror analogue tothe deflecting polarizing mirror can be installed for the polarization analysis of thereflected beam. Between the frame overlap mirror and the sample, and the sampleand the detector, various diaphragms can be placed in order to define the beam. Tofulfil the requirements of experimental stability and vibration control, all devicesare installed on the 8 m long solid optical bench, where they can be moved alongthe beam direction (see figure 1).

Pramana – J. Phys., Vol. 63, No. 1, July 2004 59

Mukul Gupta et al

Figure 2. Variation of polarization as a function of qz at various appliedmagnetic field at the polarizer at an angle of 0.5◦ (a) and variation of polar-ization at qz = 0.0036 nm−1 (b).

5. Detectors

Two 3He single detector tubes and one 3He two-dimensional multi-wire positionsensitive detector (PSD) are installed in AMOR. The EMBL 2D-PSD has an activearea of 172 × 190 mm2 with less than 2 mm spatial resolution. The obtained dataeither in monochromatic or in TOF mode are stored in ASCII or HDF5/NEXUSfile format, respectively.

6. Sample environment

The flexibility of the instrument allows for large sample environments, such ascryomagnets, evaporation chambers, furnaces or Langmuir troughs. Currently a1.8 T electromagnet with a horizontal field operational at room temperature, two



Figure 3. Variation of polarization as a function of reflected wavelengthband at an incident angle of 0.5◦.

Helmholtz coils for a small horizontal field up to 150 G applicable around largersample stages, a 2 T cryomagnet with a horizontal field and a temperature rangebetween 1.5 and 300 K, and a Langmuir trough for measurements at the air–liquidinterphase, are provided as special sample environments on AMOR. A sample cellfor measurements at the solid–liquid interface and a dedicated two-zone-heatingvacuum furnace in the temperature range of 300–1100 K are also available.

7. Polarization set-up

AMOR is operational for full polarization analysed neutron reflectometry mea-surements since June 2003. Remanent FeCoV/Ti :N supermirrors [4] are used aspolarizer/analyzer with a polarization efficiency of ∼97%. The polarization state ofthe neutron beam (either up |+〉 or down |−〉) is defined by applying a saturationfield (+350G or −350G, respectively) for a short period of time.During the setting-up of the polarizer it has been found that the polarization

of the reflected beam from the polarizer is very sensitive to: (i) the angle of thepolarizer towards the incoming beam and (ii) the guide field required to keep the po-larization. Depending on the incoming wavelength band from the choppers (tunedwith chopper speed, phase and distance between them), the polarizer has to be keptat a specific angle so as to reflect the complete incoming wavelength band. Theguide field required to keep the polarization has been determined by performingdetailed measurements of polarization by varying the magnetic field (guide field)around the polarizer. From figure 2 it can be seen that a field higher (or lower)than ∼20 G starts depolarization of the beam as the thin layers of the supermirrorstarts switching with the field. Figure 3 shows the variation of polarization as afunction of reflected wavelengths at an angle of 0.5◦. Up to a wavelength of 0.6nm polarization remains almost constant to a value of ∼97% and for the higherwavelengths the polarization decreases down to 87% for a wavelength of 0.9 nm.


Mukul Gupta et al

Figure 4. Polarized neutron reflectivity of Fe/Cr multilayer at a saturationfield of 5000 G and a remanent field of 200 G. At remanence the magneticpeak due to AF coupling can be seen clearly at double periodicity aroundqz = 0.007 nm−1. The upper curve has been shifted by a factor of 10 forclarity.

8. Recent results

Figures 4 and 5 show representative results of polarized neutron reflectivity onFe/Cr multilayers and that of liquid D2O respectively. In the case of Fe/Cr mul-tilayer, the measurements have been done at a saturation field of 5000 G and ata remanence field of 200 G. It is well-known that Fe/Cr multilayers exhibit stronganti-ferromagnetic (AF) coupling [5], when exposed to the saturation field whereall the spins are aligned ferromagnetically. As can be seen in figure 4b, the AFpeak, which is clearly observed at remanence, disappears completely at saturationfield.Figure 5 shows the reflectivity of a D2O surface measured at the Langmuir trough.

As can be seen, the scattering length density from the current measurement matcheswell with the theoretical values over four orders in dynamic range, within a reason-able time of 2–3 h for the measurement.

9. Conclusions

As can be seen from the present work, the neutron reflectometer AMOR offersextremely flexible opportunities to measure all kinds of multilayered structures,



Figure 5. Reflectivity of liquid D2O at air/water interface at room temper-ature.

magnetic or non-magnetic, to study various phenomena, e.g. adsorption, wettingand de-wetting, diffusion etc. With the available wavelength range of 0.15 nm< λ < 1.3 nm, and with TOF, q range up to 0.02 nm−1 can be obtained relativelyfast with only 2-3 angular settings. As AMOR is designed to perform reflectometrymeasurements in horizontal sample-plane geometry, it allows studying both solid–liquid and liquid–liquid interfaces. With the feasibility of selecting the chopperfrequency, width of the gating window and the chopper–detector distance a widerange of q-resolution ∆q/q = 1–10% can be selected according to experimentalrequirements.

References

[1] http://sinq.web.psi.ch/sinq/instr/amor.html[2] D Clemens, Physica B221, 507 (1996)[3] D Clemens et al, Physica B276–278, 140 (2000)[4] P Boni, D Clemens, M Senthil Kumar and C Pappas, Physica B267–268, 320 (1999)[5] A Schreyer et al, Phys. Rev. B52, 16066 (1995)


amor{thetime-of-°ightneutronre°ectometer atsinq/psi€¦ ·...

Documents